Capstone Project - The Battle of Neighborhoods

Description of the problem and the background

As a person living in Ireland, I found out about the Irish Tradition before Christmas - 12 Pubs of Christmas Pub Crawl, where you and your friends visit 12 pubs in Christmas jumpers. I suggest that this can be an extremely unique opportunity in North America. After my primary research, I also noticed that there are little similar business projects and proposed that this topic can be chosen as Capstone project

In order to do so, it is important to find out 12 best pubs according to Foursquare and calculate average distance from the mean coordinates in order to understand the most suitable city to start a business. Thus, the current information can be beneficial to potential investors.

Data descriprion

The Foursquare API will be used to extrapolate the coordinates to show it on the map. Based on the previous assignements, New York and Toronto were chosen to solve the problem. Furthermore, it became important to set limit to the needed number of pubs - 12 and to add pub category id to our query. Then, I normalised the results and got names, addresses, latitudes and longitudes of the pubs to show them on the map. The following data was ontained:

In [13]:
Out[13]:

Initial hypothesis

New York seems to be the better city to start the business based on the common sense, i.e. larger Irish diaspora and less complicated city logistics

Methodology

In order to solve the problem, it was needed, at first, to import all essential packages that were used in previous labs. Secondly, .json query was created to find out the credentials of the 12 pubs. Thirdly, folium.map was used to show the points on the maps for primary analysis (based on the algorithm of Toronto lab). As a result, the following map became obtainable for both New York and Toronto:

In [15]:
Out[15]:

Fourthly, I created the point, which contains the mean coordinates of the selected pubs, that is supposed to considered as a started point for crawl. After that, I calculated the mean distance between the starting point and the chosen pubs. Finally, this variable was the key in this "battle". The illustrative maps was then obtained for both cities using folium.map function:

In [16]:
Out[16]:

Results

The mean variable is 35% longer in Toronto compared to New York. Based on the "battle" variable, it is clear that New York is a better place for the 12 pubs crawl service.

In [18]:
Out[18]:

Discussion

The capstone project takes into account only the latitude and longtitude distance, irrespective of the means of threansport. However, 12 pubs is very seasonal thing for business. There are usually less places attended for such type of events. Looking from that perspective, i.e for crawls with less pubs, Toronto can be a better option as a starting city. Finally, no income and population effects were considered.

Conclusion

The following project illustrates potential possibility to open a business in a service sphere for pub crawls in North America. Nevertheless, the research has certain limitation and should be used only as a supporting tool for further analysis.